FBIO: Fuzzy based BIO active queue management

During past few years, Internet usage and number of applications are growing rapidly. Due to this usage growth, also amount of traffic have increased. There are different requirements along with technological improvements and new applications. To provide network throughput and quality of service to new applications has become one of the most important problems. It is necessary to design effective congestion control and queue management due to new requirements.For along time, congestion control mechanisms are using to prevent congestion in the networks. Active queue management (AQM) algorithms are the most common methods for congestion control. These work basically on managing queue size by choosing which packet to drop or not. Nowadays, many AQM algorithms exist. According to this, most of these algorithms do not work well under high traffic and different types of applications. Besides, hard to configure parameters, long delay times and most packet drops are issues to have improvements.In our thesis, existing congestion control mechanisms and active queue management algorithms are observed firstly and basic problems and solutions are investigated. This led to a new trend in using alternative techniques. Fuzzy Logic Controllers, which have ability to cope with the aforementioned problems and provide more flexibility in modeling congestion controllers as well, are most important ones. In our approach, we implement FLC into mostly knowing AQM algorithm named BIO to improve its performance. After this, our design tested and compared with traditional design.Our active queue management designs are simulated on ns-2 network simulator tool. Firstly, we choose variables that input and output for fuzzy logic controller and design membership functions of these variables. After that, IF-THEN fuzzy rules are built and associated with variables. Single-tone fuzzifier and Center Of Gravity (COG) defuzzifier methods were used in the design. As a result of measurement by simulations, it was shown that the proposed fuzzy design achieves less packet loss, short delay times, short queue sizes and high throughput compared with traditional one.